1.1 Sample Information

A brief sample information is generated from the submission table for the following analysis.

1.2 MERSCOPE Data Quality Summary

The summaries present the data quality assessment automatically generated by MERSCOPE for each experiment. We mainly focus on the transcripts level for each sample. So we’re looking for high density in transcripts, based on the transcripts count per field of view (FOV), transcript density in FOV, and frequency of transcripts detected.

Generally, log10 transcript count > 4.0 in most area can be considered as a good quality standard for human tissue.

Need to note that the low accuracy in DAPI cell boundary is not a concern, as a self-designed cell segmentation processing will take over this task.

1.2.1 NT28530x2(Normal)

1.2.2 TT28519x1(Tumor)

1.2.3 TT28519x2(Tumor)

1.2.4 NT28530x1(Normal)

1.3 Transcript Mis-Match on FOV Boundary

Due to issues with the Vizgen software, transcript alignment is incorrect during decoding, leading to misalignment problems at the field of view (FOV) boundary.

Vizgen has solved this issue by temporarily updating the decoding software to an unpublished beta version.

However, this update also made the “vzg” file being upgraded to “vzg2”. Current Vizgen Visualizer cannot recognize the new data format, which made Visualizer unusable for this dataset.

1.4 Autofluorescent issue

Autofluorescence is a major problem limiting the sensitivity of the detection of the fluorescence specifically derived from the applied dye or probe. As we can find it on DAPI imaging, there are many over-bright region.

Autofluorescence may occur due to the incomplete clearance of the tissue. Because of the presence of autofluorescence, the decoding would be confounded and the results in these cells would be artifacts.

As Dr. Tan highly recommends, we removed the portion of cell with autofluorescent issue, for better accuracy.

NT28530_DAPI_Imaging

NT28530_DAPI_Imaging_small

NT28530_dup_DAPI_Imaging

NT28530_dup_DAPI_Imaging_small

1.5 DAPI Imaging with Tumor-Associated Genes

Tumor_DAPI_Imaging

Tumor_DAPI_Imaging

Tumor_DAPI_Imaging

2.1 Cell Segmentation & Filtering

Based on the spatial information and images obtained from MERFISH, we developed a machine learning model using the Cellpose algorithm to distinguish individual cells via MERFISH DAPI images.

To ensure the data quality and accuracy of cells, we have defined the minimum and maximum values for cell volume and gene count per cell. The cell volume should be between [100, 2000], and the gene count per cell > 25.

After that, we successfully obtained high-quality and accurate cell and gene data. We are now ready to perform statistical analysis.

For analysis platform, all of the analysis and visualization are performed using Scanpy, a commonly used Python package toolkit for analyzing single-cell gene expression data.

2.1.2 Transcript Count Violin

The transcript level is a basic aspect of data quality control. After filtering, we have ensured that the minimum transcript count per cell > 25.

Transcript Count Violin After Filtering

2.2 Batch Effect & Dimension Reduction

We conducted Leiden UMAP clustering, a non-linear, manifold-aware dimension reduction algorithm, on our dataset to subset the cells into clusters. The following annotations are based on the umap clustering.

Umap of cells and colored by batch

3.1 Markers for type annotation

3.2 Cell Type Umap

3.3 Cell Type Umap (group)

3.4 Cell Type Spatial Map

With the annotations, we can map cell types back to their physical positions and create a spatial map for each sample.

3.5 Cancer & T cell & B cell Spatial Map

3.6 Cell Type Qualification Analysis

3.7 Cell Type Proportion Analysis

Based on the quantification of individual cell types, we are capable to compare the differences in cell numbers between various samples and genotypes. The proportion of each cell type is calculated by dividing the count of that specific cell type by the total cell count. The results are then visualized in a bar plot.

Notably, B cells, mast cells, and T cells show an obvious increase in tumor samples.

4.1 all

4.2 Cancer cell

4.3 Dendritic

4.4 Endothelial

4.5 Fibroblast

4.6 Macrophage

4.7 Mast cell

4.8 Smooth Muscle cell

4.9 T cell

5.1 all

5.2 Dendritic

5.3 Endothelial

5.4 Fibroblast

5.5 Macrophage

5.6 Mast cell

5.7 Smooth Muscle cell